JP3343919B2 - Mask, circuit element manufacturing method and exposure method - Google Patents

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a circuit for a semiconductor device or the like.
A mask used in an exposure device that transfers a pattern onto a substrate
And a method for manufacturing a circuit element using the mask, and an exposure method.

[0002]

2. Description of the Related Art For forming a circuit pattern of a semiconductor element or the like,
In general, a process called a photolithographic technique is required. In this step, a method of transferring a reticle (mask) pattern onto a sample substrate such as a semiconductor wafer is usually adopted. A photosensitive photoresist is applied on the sample substrate, and a circuit pattern is transferred to the photoresist according to an irradiation light image, that is, a pattern shape of a transparent portion of the reticle pattern. In a projection type exposure apparatus (for example, a stepper), an image of a circuit pattern to be transferred drawn on a reticle is projected and formed on a sample substrate (wafer) via a projection optical system.

FIG. 14 shows a schematic configuration of a conventional projection type exposure apparatus (stepper) as described above. In the conventional projection type exposure apparatus, the light flux passing through the Fourier transform surface with respect to the reticle pattern 12 is made substantially uniform within a substantially circular shape (or within a rectangular shape) centered on the optical axis of the illumination optical system. The illumination light beam L0 is converted into an illumination light beam L10 having a predetermined shape by an aperture stop (spatial filter) 15b in the illumination optical system.
The illumination light beam L10 irradiates the pattern 12 of the reticle 11 via the condenser lens 8. Here, the spatial filter 15b is arranged on the Fourier transform plane 15 (hereinafter abbreviated as the illumination system pupil plane 15) for the reticle pattern 12 or in the vicinity thereof, and has a substantially circular shape centered on the optical axis AX of the projection optical system 13. At the opening of the area,
A secondary light source (surface light source) image formed in the pupil plane is limited to a circular shape. Thus, the reticle pattern 12 is imaged on the resist layer of the wafer 19 via the projection optical system 13 by the illumination light having passed through the reticle pattern 12. Here, the solid line representing the light beam represents the center of the light beam emitted from one point. At this time, the ratio between the numerical aperture of the illumination optical system (15b, 8) and the reticle-side numerical aperture of the projection optical system 13, that is, the so-called σ value, is determined by the aperture stop (for example, the aperture diameter of the spatial filter 15b). Generally, about 0.5 to 0.6 is used. A conventional mask pattern has the same shape as a pattern to be formed on a wafer as a semiconductor integrated circuit pattern, for example.
Or it was a similar pattern. As the mask, for the same-size exposure (contact method, proximity method, mirror projection method, etc.), use a mask having the same shape (combination) as the pattern to be formed on the wafer, and use reduced projection exposure (stepper method). Etc.), the mask pattern uses a pattern larger than the pattern to be formed on the wafer by the reduced amount. If the reduction ratio is 1/5, the dimension on the reticle can be obtained by making it five times the dimension on the wafer (converted on the mask side).

Incidentally, the σ value (coherence factor) of the illumination type of a conventionally used projection exposure apparatus is σ
= 0.5 to 0.6, and the coherence on the reticle was low. Therefore, light interference between adjacent patterns was not a problem. Now, the illumination light L10 is diffracted by the patterned pattern 12 on the reticle 11, from the pattern 12 0-order diffracted light D _0, + 1-order diffracted light D _P, and -1-order diffracted light D _m is generated. The respective diffracted lights (D ₀ , D _m , D _P ) are condensed by the projection optical system 13 and generate interference fringes on the wafer 19. This interference fringe is an image of the pattern 12. At this time, the zero-order diffracted light D
_The angle θ (reticle side) between ₀ and the ± first-order diffracted lights D _P and D _m is determined by sin θ = λ / P (λ: exposure wavelength, P: pattern pitch).

When the pattern pitch becomes finer, sin θ increases. When sin θ becomes larger than the reticle-side numerical aperture (NA _R ) of the projection optical system 13, the ± 1st-order diffracted lights D _P and D _m pass through the projection optical system 13. It cannot be transmitted.
At this time, only the _zero-order diffracted light D ₀ reaches the wafer 19, and no interference fringes occur. That is, when sin θ> NA _R , the image of the pattern 12 cannot be obtained, and the pattern 1
2 cannot be transferred onto the wafer 19.

From the above, in the conventional projection type exposure apparatus, the pitch P which satisfies sin θ = λ / P ≒ NA _R
Was given by: P ≒ λ / NA _R (1) From this, since the minimum pattern size is half of the pitch P, the minimum pattern size is about 0.5 · λ / NA _R , but in the actual photolithography process, Some degree of depth of focus is required due to curvature, the effects of process steps, etc., or the thickness of the photoresist itself. Therefore, the practical minimum resolution pattern size is expressed as k · λ / NA. here,
k is called a process coefficient and is about 0.6 to 0.8.
Since the ratio between the reticle-side numerical aperture NA _R and the wafer-side numerical aperture NA _W is the same as the imaging magnification of the projection optical system, the minimum resolution pattern size on the reticle is k · λ / NA _R ,
The minimum pattern size on the wafer is k · λ / NA _W = k
· Λ / B · NA _R (where B is the imaging magnification (reduction ratio)) becomes.

Therefore, in order to transfer a finer pattern, it was necessary to select whether to use an exposure light source having a shorter wavelength or to use a projection optical system having a larger numerical aperture. Of course, efforts can be made to optimize both the exposure wavelength and the numerical aperture. Also, a so-called phase shift reticle that shifts the phase of the transmitted light from a specific portion of the transmitted portion of the circuit pattern of the reticle by π from the phase of the transmitted light from another transmitted portion, for example, is disclosed in
50811. Use of this phase shift reticle enables transfer of a finer pattern than in the past.

[0008]

However, in the conventional projection type exposure apparatus as described above, shortening the wavelength of the illumination light source (for example, 200 nm or less) from the present time requires an appropriate optical element usable as a transmission optical member. It is difficult at present because of the absence of materials. Also, the numerical aperture of the projection optical system is already close to the theoretical limit at present, and it is almost impossible to increase the numerical aperture further.

Further, even if the aperture can be made larger than the current situation, the depth of focus represented by ± λ / 2NA ² rapidly decreases with an increase in the numerical aperture (NA), and it is actually used.
There is a problem that the depth of focus required for (pattern transfer) is further reduced. On the other hand, the phase shift reticle has many problems such as a high manufacturing cost due to its complicated manufacturing process, and no inspection and repair method has been established yet.

[0010] The present invention has been made in view of the above problems, and can improve the transfer accuracy of a fine pattern.
Reticles, especially those with high resolution and large depth of focus
A reticle suitable for use with optical technology, for example, to obtain a reticle consisting of only a light-transmitting portion and a light-shielding portion that is easier to manufacture, inspect, and modify than a phase shift reticle. And In addition, using the reticle,
With the object of providing a circuit element manufacturing method for forming a throat
I do. It is still another object of the present invention to provide an exposure method that can obtain a high resolution and a large depth of focus even when a reticle including only a light-transmitting portion and a light-shielding portion is used.

[0011]

According to one aspect of the present invention, the edge of a pattern on a mask is reduced by a very small amount depending on whether the edge is isolated or not. The line width of the pattern is modified so as to expand or contract only. A projection type exposure apparatus used in an exposure method according to another aspect of the present invention is configured in principle as shown in FIG. 13, the same members as those in FIG. 14 are denoted by the same reference numerals. In FIG. 13, the illumination light beam L0 irradiates the pattern 12 of the reticle 11 via the spatial filter 15a and the condenser lens 8. Here, the spatial filter 15a is arranged at or near the pupil plane of the illumination system, and the transmission portion of the spatial filter 15a has an illumination optical system or an illumination optical system corresponding to the fineness of the pattern 12 on the reticle 11 and the periodic direction. It is provided at a position decentered from the optical axis AX of the projection optical system. Therefore,
The illumination light beam L0 is limited to an illumination light beam L9 that passes through a local region having a center at a position decentered from the optical axis AX of the illumination optical system or the projection optical system in a pupil plane of the illumination system or a plane near the pupil plane. The position at which the illumination light beam L9 is transmitted in the illumination system pupil plane or a plane near the illumination system pupil surface determines the incident angle and direction of the light beam L9 incident on the reticle 11. Therefore, in the projection exposure method of the present invention, the illumination system pupil is used. The incident angle 照明 of the illumination light beam L9 incident on the reticle 11 according to the position where the illumination light beam L9 passes through the surface or a surface in the vicinity thereof.
And the direction can be controlled almost arbitrarily. Also,
For the above purpose, in the exposure method according to the present invention, the line width of the pattern is expanded or reduced by a very small amount depending on whether the edge of the pattern is isolated or at the end. The modified mask was used.

[0012]

The principle of the exposure method according to the present invention will be briefly described. In FIG. 13, a circuit pattern 12 drawn on a reticle generally includes many periodic patterns. Accordingly, from the illumination light beam reticle pattern 12 L9 is irradiated, 0-order diffracted light component D ₀ and ± 1-order diffracted light component D _P, is D _m and more diffracted light components of higher order, depending on the fineness of the pattern direction Occurs. At this time, since the illumination light beam (center line) is incident on the reticle 11 at an inclined angle, the generated diffracted light components of each order also have a certain inclination (angle shift) as compared with the case where the illumination is performed vertically. From 12 The illumination light L9 in FIG. 13 is incident on the reticle 11 at an angle of ψ with respect to the optical axis.

The illumination light L 9 is generated by the reticle pattern 12.
Diffracted and proceeds in a direction inclined by ψ with respect to the optical axis AX 0
Next order diffracted light D₀, Θ for the 0th order diffracted light_PJust lean forward
+ 1st order diffracted light D_P, And zero-order diffracted light D₀For θ_mIs
-1st order diffracted light D _mOccurs. Where
The bright light L9 is light of the projection optical system 13 which is telecentric on both sides.
Enter the reticle pattern at an angle ψ with respect to the axis AX
The zero-order diffracted light D₀Also the optical axis A of the projection optical system
It proceeds in a direction inclined by an angle ψ with respect to X.

Therefore, the + 1st-order diffracted light D _P travels in the direction of (θ _P + に 対 し て) with respect to the optical axis AX, and the −1st-order diffracted light D _m in the direction of (θ _m −ψ) with respect to the optical axis AX. proceed. At this time, the diffraction angles θ _P and θ _m are sin (θ _P + ψ) −sinψ = λ / P (2) sin (θ _m −ψ) + sinψ = λ / P (3) Here, the + 1st-order diffracted light _Dp and the -1st-order diffracted light D
It is assumed that both _m are transmitted through the pupil plane 18 of the projection optical system 13 (the Fourier transform plane of the reticle pattern 12 in the projection optical system 13).

When the diffraction angle increases as the reticle pattern 12 becomes finer, first-order diffracted light D _P traveling in the direction of the angle (θ _P + ψ) cannot pass through the pupil 18 of the projection optical system 13. That is, sin (θ _P + ψ)> NA _R
It comes to the relationship. However, the illumination light L9 is because the incident inclined with respect to the optical axis AX, -1-order diffracted light D _m at a diffraction angle in this case is permeable to the projection optical system 13. That is, a relationship of sin (θ _m −ψ) <NA _R is established.

Therefore, the zero-order diffracted light D ₀ is placed on the wafer 19.
When -1 interference fringes caused by two light beams of the diffracted light D _m occurs.
This interference fringe is an image of the reticle pattern 12. When the reticle pattern 12 has a line-and-space ratio of 1: 1, the image of the reticle pattern 12 is formed on a resist layer applied on the wafer 19 with a contrast of about 90%. Can be patterned.

The resolution limit at this time is when sin (θ _m -ψ) = NA _R (4), and therefore, NA _R + sinψ = λ / PP = λ / (NA _R + sinψ) (5) ) Is the pitch of the smallest transferable pattern on the reticle side.

As an example, if sinψ is determined to be about 0.5 × NA _R , the minimum pitch of the pattern on the reticle that can be transferred is P = λ / (NA _R + 0.5NA _R ) = 2λ / 3NA _R ( 6)

On the other hand, when a conventional projection type exposure apparatus is used as described above, as shown in FIG.
The above light amount distribution was within a circular area centered on the optical axis AX of the projection optical system 13, and its resolution limit was P ≒ λ / NA _R as shown in the equation (1). Therefore, as is apparent from comparison with the equation (6), it can be seen that a higher resolution can be realized than the conventional projection type exposure apparatus.

Next, by exposing the reticle pattern to exposure light at a specific incident angle, a focus is formed by a method of forming an image pattern on a wafer using the 0th-order diffracted light component and the 1st-order diffracted light component. The reason why the depth is increased will be described. As shown in FIG. 14, the wafer 19 is
, The diffracted light components that leave one point in the reticle pattern 12 and reach one point on the wafer 19 pass through any part of the projection optical system 13. All have the same optical path length. For this reason, even when the 0th-order diffracted light component passes almost the center (near the optical axis) of the pupil plane 18 of the projection optical system 13 as in the related art,
The 0th-order diffracted light component and the other diffracted light components have the same optical path length, and the mutual wavefront aberration is also zero. However, the wafer 19
Is in the defocus state where the focal position of the projection optical system 13 does not match, the optical path length of the obliquely incident high-order diffracted light component is in front of the 0th-order diffracted light component passing near the optical axis (projection optics). The distance is shorter at a position away from the system 13) and longer at a position behind the focal point (a direction closer to the projection optical system 13), and the difference depends on the difference in the incident angle. Therefore, 0 order, ± 1
Next, each of the diffracted light components forms a wavefront aberration, and blurs before and after the focal position.

The wavefront aberration due to the above-mentioned defocus is represented by ΔF, the amount of deviation from the focal position of the wafer 19, and the sine of the incident angle θ _w when each diffracted light component is incident on the negative side.
When (r = sinθ _w), it is an amount given by ΔFr ^2/2. At this time, r represents the distance of each diffracted light component from the optical axis AX on the pupil plane 18. In the conventional projection exposure apparatus shown in FIG. 14 (stepper), 0-order diffracted light D ₀ becomes r (0-order) = 0 because passing near the optical axis AX. On the other hand, ± first-order diffracted lights D _P and D _m are r (first order) = M · λ /
P (M is the imaging magnification of the projection optical system).

Accordingly, the 0th-order diffracted light D ₀ and the ± 1st-order diffracted light D
_P, the wavefront aberration due to defocus of the D _m is, ΔF · M
^{2 (λ} / P) becomes a ^2/2. On the other hand, in the projection type exposure apparatus used in the exposure method according to the present invention, the 0th-order diffracted light component D ₀ is generated in a direction inclined from the optical axis AX by an angle ψ as shown in FIG. The distance of the folded light component from the optical axis AX is r (0th order) = M · sinψ.

Furthermore, the distance from the optical axis in the -1 pupil plane of the diffracted light component D _m is r (-1 order) = M · sin (θ _m
−ψ). At this time, sinψ = sin (θ
if the _{m -ψ),} 0 relative wavefront aberration becomes zero due to defocusing of the order diffracted light component D ₀ and -1-order diffracted light component D _m, even the wafer 19 is slightly deviated in the optical axis direction from the focal position pattern Twelve image blurs do not occur as much as before. That is, the depth of focus increases. Also, as in equation (3), sin (θ _m −ψ) + sinψ
= Λ / P, the reticle 11 of the illumination light beam L9
Angle ψ to the pattern of pitch P, sin
If the relationship is defined as ψ = λ / 2P, it is possible to greatly increase the depth of focus.

In the exposure method according to the present invention,
It has been experimentally confirmed that the depth of focus of the isolated pattern can be larger than that of the conventional method. Next, the reason for correcting the peripheral portion of the circuit pattern will be described by taking a periodic pattern as an example. Illuminating the periodic pattern with the illumination method of illuminating the reticle at the oblique incident angle described above,
When the pattern is exposed through the projection optical system, a very good pattern image as designed is obtained for a relatively inner portion of the periodic pattern. However, at the end portion of the periodic pattern (the end portion in the direction perpendicular to the periodic direction in each pattern element of the periodic pattern), the formed pattern image is deformed by tapering, film thinning, and the like,
At both ends of the periodic pattern (both ends of the periodic pattern element at both ends in the periodic direction), the line width of the formed pattern image is thin and deformed.

FIG. 12 shows an example of the deformation of the periodic pattern. FIG. 12A shows a light-shielding portion (shaded portion).
FIG. 12E illustrates a one-dimensional periodic pattern including three light-transmitting portion pattern elements provided on the ground, and FIG. 12E illustrates a primary pattern including three light-shielding (hatched portion) pattern elements provided on the light-transmitting portion ground. Represents the original periodic pattern. 12A and 12E.
In both the patterns shown in (1) and (2), the desired reticle-side converted value (multiplied by the projection magnification) of the desired resist image shape after exposure transfer of the pattern onto the wafer is directly used as the pattern. Hereinafter, the reticle-side converted value of this desirable resist image is referred to as “design value”. FIG. 12 (A),
The pattern shown in (E) is a line and space pattern with a duty of 1: 1 arranged at a pitch 2a. FIG.
2 (B) and (F) show the light quantity distribution (optical image) on the wafer surface in the M1-M2 cross section in the patterns shown in FIGS. 12 (A) and 12 (E), respectively. FIG. 12 (C),
(G) shows the optical image on the wafer surface in the section M3-M4 in the patterns shown in FIGS. 12 (A) and 12 (E), respectively. Note that the broken line Eth1 in FIGS. 12B and 12C is
The exposure amount required to completely remove the negative resist is shown, and the broken line Eth2 in FIGS. 12F and 12G indicates the exposure amount necessary to completely remove the positive resist.

FIG. 12D shows a top view of a negative resist image when the pattern shown in FIG. 12A is exposed on the negative resist, and FIG. 12H shows the pattern shown in FIG. FIG. 3 is a perspective view of a positive resist image when is exposed on a positive resist. The lighting conditions at this time are shown in FIG.
(A) and (E) are optimized for the pattern, and the line width a is a fine pattern transferred onto the wafer using a projection exposure apparatus employing a special illumination method as shown in FIG. Is about the resolution limit. For example, in the case of forming a line leaving pattern using a negative resist, as shown in FIG.
For this reason, as shown in FIG.
Tapering occurs at the longitudinal end.

Further, as shown in FIG.
Since the intensity of the two patterns at both ends of the three patterns shown in FIG. 12A is reduced, the line width of the negative resist image shown in FIG. I will. On the other hand, when a line-remaining pattern is formed using a positive resist, as shown in FIG. 12F, the optical image in the longitudinal direction of the pattern has a drooping change in the light amount distribution at the end portion, and sufficient light shielding is not achieved. become unable. For this reason, in the positive resist image shown in FIG. 12H, the film thickness is reduced at the end portion in the longitudinal direction. Is performed), resulting in a trapezoidal resist image. Further, since the light shielding by the two light-shielding patterns at both ends of the three light-shielding portion patterns shown in FIG. 12E is not sufficient, the line width at both ends is smaller than that at the center. In FIG. 12, a three-line periodic pattern is used as an example. However, even with an arbitrary periodic pattern, an image is generated between the inner part and the inner part of the periodic pattern and both ends. Has a smaller line width. This is because, other than both ends (inner part) of the periodic pattern, other patterns always exist near the inner part of the pattern (pattern edge) (underlying area where the distance from the edge is within the minimum pitch). This is because other patterns do not exist near the patterns at both ends (pattern edges at both ends). In addition, tapering and film reduction at the terminal end also occur.

In order to prevent the above-mentioned deformation, the line width at the terminal end portion is increased by a minute amount so as to correct the taper, the decrease in film thickness, and the line width at the terminal end portion. A mask whose pattern is corrected so as to increase the line width by a small amount so as to correct the width reduction is used. On the contrary, the correction is performed in such a manner that the line width at the center portion is reduced by a small amount without changing the line width at the terminal portion, or the isolated edge portion (an edge where no other pattern exists near the edge). The same effect can be obtained by performing correction so as to reduce the line width by a small amount in portions other than the isolated edge portion without performing the correction in (Part).

In the above description, the periodic pattern has been described as an example. However, even in an isolated pattern, a taper or thinning occurs at the end portion, and a line width decreases at the center portion. This also occurs because there is no other pattern in the vicinity of the isolated edge portion, particularly the end portion thereof, as described above. In order to cope with a case where an isolated pattern is present in the periodic pattern, a corrected mask is used for the isolated pattern.

As described above, whether the edge of a pattern is isolated (when the same type of pattern element does not exist near the pattern) or close (when the same type of pattern element exists near the pattern) Accordingly, the line width is corrected so as to increase or decrease by a small amount depending on whether it is the end portion or the center portion of the pattern. According to the present invention described above, even if a conventional reticle consisting only of a light-shielding portion and a transmission portion is used, the resolution and the depth of focus can be reduced simply by correcting a part of the shape and illuminating the reticle with a predetermined inclination. Both are greatly improved, and good image quality can be obtained in all circuit patterns such as periodic patterns and isolated patterns.

[0031]

FIG. 1 is a diagram showing a schematic configuration of a projection type exposure apparatus most suitable for applying an exposure method according to an embodiment of the present invention. The light beam generated by the light source 1 is reflected by the elliptical mirror 2 and the reflecting mirror 3, and passes through a lens system 4 to a fly-eye lens 5.
Incident on. The light beam emitted from the fly-eye lens 5 enters a light splitter 7 such as an optical fiber via a lens system 6. The light splitter 7 divides the light beam incident from the incident portion 7i into a plurality of light beams and emits the light beams from the plurality of emission portions 7a and 7b. The emission surfaces of the emission units 7a and 7b are formed by the pattern 12 on the reticle 11.
Lens system 8, 10 and reflecting mirror 9 for the surface where
Plane (pupil plane of illumination optical system) 1 which becomes a Fourier plane through
5 or in the vicinity thereof.

The optical axis AX at the position of these emission parts 7a, 7b
Is determined according to the angle of incidence of the illumination light beam on the reticle 11. The plurality of light beams emitted from the emission units 7a and 7b are transmitted to the lens system 8, the reflecting mirror 9, and the lens system 1
The reticle 11 is illuminated at a predetermined angle of incidence through each of the reticle 11. The reticle 11 is mounted on a reticle stage RS. The diffracted light generated by the pattern 12 on the reticle 11 forms an image on the wafer 14 via the projection optical system 13 and transfers the image of the pattern 12. The wafer 14 is mounted on a wafer stage WS that can move in a two-dimensional direction in a plane perpendicular to the optical axis AX, and can sequentially move the transfer area of the pattern 12. In addition, this exposure apparatus
A shutter for controlling the amount of light applied to the reticle 11,
And an irradiance meter and the like in the illumination optical system.

As the light source 1, a bright line lamp such as a mercury lamp or a laser light source is used. Further, in this embodiment, an optical fiber is used as the light splitter 7 for splitting the illumination light beam, but another member, for example, a diffraction grating or a polygon prism may be used. In the above configuration, the exit surface of the light source 1 and the fly-eye lens 5 (a surface substantially conjugate to the pupil surface 15 of the illumination optical system), the exit surface of the optical fiber 7 (the pupil surface 15), and the pupil surface 18 of the projection optical system 13 Conjugate to each other, and
The incident surface of the fly-eye lens 5 and the incident surface of the optical fiber 7, the pattern surface of the reticle 11, and the transfer surface of the wafer 14 are conjugate to each other.

In addition, the reticle 11 side from the light splitter 7,
That is, another fly-eye lens may be added near the exit surfaces of the exit portions 7a and 7b for uniform illumination. At this time, the fly-eye lens may be a single fly-eye lens, or may be configured by a plurality of fly-eye lens groups provided corresponding to the respective exit ends. Further, depending on the correction state of the chromatic aberration of the projection optical system 13 and the illumination optical systems 1 to 10, a wavelength selection element (such as an interference filter) may be added to the illumination optical system.

When the reticle 11 is illuminated by using the above-described apparatus, the illuminating light beams emitted from the emission portions 7a and 7b of the light splitter 7 enter the reticle 11 at a predetermined angle as described in the operation section. Therefore, it is possible to pass two luminous fluxes together with any one of the ± first-order diffracted light generated from the reticle pattern and the zero-order diffracted light through the pupil plane 18 of the projection optical system, and to reduce the pitch ( It is possible to resolve even a (fine) pattern. Further, in the exposure method according to the present invention, the reticle pattern is illuminated with an incident direction and an incident angle corresponding to the period (pitch) of the periodic pattern and the periodic direction (described later in detail). For this reason, the injection units 7a, 7b
Is desirably movable in the plane of the pupil plane 15, as shown in FIGS.

FIG. 2 is a cross-sectional view of the emission section as viewed from a direction perpendicular to the optical axis, and FIG. 3 is a plan view as viewed from the optical axis direction.
Here, four fiber exit ends 7a, 7b, 7c and 7d are used as means for creating an arbitrary light amount distribution on the pupil plane, and each fiber exit end is discretely decentered from the optical axis AX. Position and at substantially the same distance from the optical axis AX. 2 and 3, the fiber emitting ends 7a to 7d are connected to support rods 17a, 17b, and 17 respectively.
The movable members 16a, 16b, 16c,
16d makes it possible to expand and contract in the direction perpendicular to the optical axis. The movable members 16a to 16d themselves are also fixed guides 1.
It is movable in the circumferential direction around the optical axis along 6e.

Further, the dividing member 7 is replaceable with another dividing member, and the dividing member may be exchanged and used according to the reticle pattern. Next, the optimal position of the illumination light beam on the pupil plane with respect to the periodic pattern will be described with reference to FIGS. FIG. 4 is a schematic diagram showing the illumination system pupil plane 15 to the projection optical system 13 of the exposure apparatus in FIG. 1. The pupil plane 15 and the reticle pattern 12 are in a Fourier transform relationship by the condenser lens 8. I have. Pupil plane 1
5, the distributions 7c1 and 7d1 of the illumination light fluxes correspond to the optical axis AX.
Are distributed around different positions. The focal length of the condenser lens 8 is f, and the distance from the pupil surface 15 to the condenser lens 8 and the distance from the condenser lens 8 to the reticle pattern surface 12 are both F.

FIGS. 5A and 5C are diagrams each showing an example of a partial pattern formed in the reticle pattern 12, and FIG. 5B is a view showing the case of the reticle pattern shown in FIG. Fourier transform plane 15 at each center of optimal illumination light flux
FIG. 5D shows the position on (pupil plane 15), and FIG.
FIG. 7 is a diagram showing the position on the Fourier transform plane 15 of each center of the illumination light beam that is optimal for the reticle pattern of FIG. FIG.
(A) is a so-called one-dimensional line-and-space pattern in which transmissive portions and light-shielding portions are arranged in a band shape in the Y direction with the same width, and they are regularly arranged at a pitch P in the X direction. Here, the light-shielding portions are indicated by hatched portions. At this time, the optimum position of each illumination light beam is on a line segment Lα in the Y direction assumed in the Fourier transform plane as shown in FIG.
And any position on the line segment Lβ. FIG. 5B shows a Fourier transform plane (pupil plane) 15 for the reticle pattern 12.
Are viewed from the optical axis AX direction, and the coordinate systems X and Y in the plane 15 are the same as FIG. 5A in which the reticle pattern 12 is viewed from the same direction.

In FIG. 5B, the distances α and β from the center C through which the optical axis AX passes to the line segments Lα and Lβ are α = β.
β, where λ is the exposure wavelength, α = β = f · (1
/ 2) · (λ / P). The distances α and β are expressed as f · s
When expressed as inψ, sinψ = λ / 2P, which is consistent with the numerical value described in the section of the operation. Therefore, the injection unit 7
If the center positions of a and 7b are on the line segments Lα and Lβ, the 0th-order diffracted light generated by the illumination light from each illumination light beam and ± 1 with respect to the line and space pattern as shown in FIG. The two diffracted lights with either one of the next-order diffracted lights pass through the projection optical system pupil plane 15 at a position that is substantially equidistant from the optical axis AX. Therefore, as described above, the depth of focus for the line and space pattern (FIG. 5A) can be maximized, and high resolution can be obtained.

Next, FIG. 5C shows a reticle pattern.
The pattern is a two-dimensional periodic pattern, and the pattern pitch in the X direction (horizontal direction) is Px, and the pitch in the Y direction (vertical direction)
The pitch is Py. FIG. 5D is a diagram showing an optimum position of each illumination light beam in this case, and the position and rotation relationship with FIG. 5C is the same as the relationship in FIGS. 5A and 5B. When illumination light is incident on a two-dimensional pattern as shown in FIG. 5C, the periodicity of the pattern in the two-dimensional direction (X: Px,
Y: Py), a diffracted light is generated in a two-dimensional direction. Also in the two-dimensional pattern as shown in FIG. 5C, one of the 0th-order diffracted light and ± 1st-order diffracted light in the diffracted light is substantially equidistant from the optical axis AX on the pupil plane 15 of the projection optical system. In this case, the depth of focus can be maximized.
Since the pitch in the X direction is Px in the pattern of FIG. 5C, α = β = f · (１ ／) as shown in FIG.
If the center of the illumination light beam is on the line segments Lα and Lβ that are (λ / Px), the depth of focus can be maximized for the X-direction component of the pattern. Similarly, γ = ε = f · (1 /
2) If the center of the illuminating light beam is on the line segments Lγ and Lε that become (λ / Py), the depth of focus can be maximized for the pattern Y-direction component.

The reticle pattern 12 is shown in FIG.
In the case where a two-dimensional periodic pattern is included as shown in FIG. 1, when focusing on one specific zero-order diffracted light component, X on the pupil plane 18 of the projection optical system is centered on the one zero-order diffracted light component.
There may be a first or higher order diffracted light component distributed in the direction (first direction) and a first or higher order diffracted light component distributed in the Y direction (second direction). Therefore, assuming that a two-dimensional pattern is favorably imaged with respect to one specific 0th-order diffracted light component, one of the higher-order diffracted light components distributed in the first direction and the higher-order diffracted light component distributed in the second direction are used. The position of the specific zero-order diffracted light component is changed to the position of one illumination light flux so that three of the folded light component and the specific zero-order diffracted light component are distributed on the pupil plane 18 at substantially the same distance from the optical axis AX. You can adjust the correspondence. For example, in FIG. 5D, it is preferable that the center position of each illumination light beam coincides with any one of the points Pζ, Pη, Pκ, and Pμ. Each of the points Pζ, Pη, Pκ, and Pμ is a line segment Lα or Lβ (the position optimal for the periodicity in the X direction, ie, the 0th-order diffracted light and one of the ± 1st-order diffracted lights in the X direction are on the projection optical system pupil plane 18). And the line segments Lγ and Lε (optimal positions for the periodicity in the Y direction).
Are the optimal light source positions in both the X direction and the Y direction.

As described above, when the illuminating light from each illuminating light flux arranged at each position shown in FIG. 5B or 5D enters the reticle pattern 12, the 0th-order light diffracted light component _DO and +1
One of the first-order diffracted light component D _{R and the} minus first-order diffracted light component D _m is transmitted along the optical axis AX on the pupil plane 18 in the projection optical system 13.
Through an optical path that is approximately equidistant from Therefore, as described in the operation section, a projection type exposure apparatus with high resolution and large depth of focus can be realized.

As described above, the reticle pattern 12 shown in FIG.
Although only the two examples shown in (A) or (C) were considered, even for other patterns, attention was paid to the periodicity (fineness), and a + 1st-order or -1st-order diffracted light component from that pattern was used. And two luminous fluxes of the 0th-order diffracted light component,
In the pupil plane 18 in the projection optical system, the center of each illumination light beam (the center of the emission portions 7a and 7b) may be arranged at a position passing through an optical path that is substantially equidistant from the optical axis AX. FIG.
In the pattern examples (A) and (C), since the ratio (duty ratio) of the line portion to the space portion is 1: 1, ± 1st-order diffracted light is strong in the generated diffracted light. For this reason, attention has been paid to the positional relationship between one of the ± 1st-order diffracted lights and the 0th-order diffracted light. However, when the pattern is different from the duty ratio of 1: 1 or the like, the other diffracted lights, for example, one of the ± 2nd-order diffracted lights The positional relationship between the zero-order diffracted light and the 0th-order diffracted light may be substantially equidistant from the optical axis AX on the projection optical system pupil plane 18.

In the above description, a pattern having a two-dimensional direction is assumed at the same location on the reticle as a two-dimensional pattern. However, when a plurality of patterns having different directions exist at different positions in the same reticle pattern. The above method can be applied also to the above. When the pattern on the reticle has a plurality of directions or finenesses, the optimal position of each illumination light beam corresponds to each directionality and fineness of the pattern as described above. Each illumination light beam may be arranged at an average position. In addition, the average position may be a load average in which a weight according to the fineness and importance of the pattern is added.

By determining the position of the illumination light beam in the pupil plane of the illumination optical system as described above, the maximum depth of focus can be obtained for each of various reticle patterns. The pitch and directionality of the pattern are substantially common to several types of semiconductor integrated circuits. For example, in the case of a memory element or the like, 1 Mbit,
There are types such as 4 Mbits and 16 Mbits depending on the degree of integration, and some of them are almost common. Furthermore, the pattern pitch is subdivided by differences in access time and word composition. At a certain time (about several months to one year), the pattern pitch of memory elements manufactured on one production line (including a projection exposure apparatus) At present, it is almost constant regardless of these types. Therefore, in practice, as described above, even if the illumination system is not optimized for each reticle pattern, at least a part of the pitch and directionality of the reticle pattern used at a certain time are averaged to improve the depth of focus. Even if the lighting system is fixed as described above, there is almost no problem. The slightest deviation from the best condition for each reticle in the illumination system reduces the depth of focus slightly from the best case, but still is still large enough.

It is preferable that each illumination light beam is symmetrically arranged in the pupil plane 15 with respect to the optical axis AX. This is to prevent the reticle pattern projection image from laterally shifting in the in-plane direction of the wafer (so-called telecentric shift) even when the wafer 19 is slightly defocused. Next, the reticle pattern used in the exposure method according to the present invention will be described.

The pattern shown in FIG. 6A has linear pattern elements (P1, P2, P3, P
4) is a line and space pattern (periodic pattern) in which four lines are arranged in the X direction. Linear pattern elements P1 to P
The line width and spacing of No. 4 are both a. Here, in order to simplify the display of the dimensions, all the dimensions are on the reticle side. That is, it is assumed that the size of the resist image is enlarged by the magnification (the reduction ratio). When the pattern of FIG. 6A is projected and exposed on the wafer coated with the negative resist using the exposure apparatus of FIG. 1 employing the above-described illumination method, as described in the operation section, the pattern of FIG. As described above, a resist image (negative resist) is formed in which the taper occurs at the terminal end (the end in the longitudinal direction) and the line width at both ends is reduced. Therefore, the reticle pattern is corrected as shown in FIG.

First, correction of the line width of the pattern elements P1 and P4 at both ends will be described. The resist image to be formed on the wafer is as shown in FIG. 6D, which is used as design data (data of circuit design). The line width in the X direction of the pattern elements P1 and P4 at both ends in FIG. 6C is increased by b to a + b. In the method of thickening, isolated edges (edges having no other patterns in the vicinity) E1 and E8 of the pattern elements P1 and P4 at both ends are extended by b in the X direction.

Next, the modification of the terminal portion will be described. The line width in the X direction at the end of the pattern element in the longitudinal direction (Y direction) is increased by c to a + c. At this time, at the end portions of the pattern elements P1 and P4, since both of the above corrections are performed, the line width becomes a + b + c and becomes wider by b + c. Accordingly, the end of each edge (E3, E4) and (E5, E6) of the pattern elements P2 and P3 on the inner side is X-shaped.
Extend in the direction by 0.5c. Then, respective edges (E1, E8) outside the pattern elements P1, P4 at both ends.
Are extended in the X direction by b + 0.5c, and the ends of the inner edges (E2, E7) are extended by 0.5c in the X direction.
Expand by one. The length in the Y direction in which correction is performed at the end portion is d. Hereinafter, the “end portion of the pattern element” refers to a portion of the pattern element having a length of about d in the longitudinal direction.

Although the pattern elements P1 and P4 lose their symmetry before and after the correction, the pattern elements P1 to P4
Is a single periodic pattern, and the symmetry is maintained when the periodic pattern is viewed as a whole. Well, each a,
The actual dimensions of b, c, and d will be described. “a” is a desired dimension of the resist image (dimension according to the data of the circuit design), and may be an arbitrary value. The actual dimensions of b, c, and d depend on the illumination condition of the exposure apparatus used in the present invention (illumination optical system pupil plane 1).
5, the pitch Pm (minimum pattern size) of the periodic pattern which is the resolution limit at the position of the illumination light beam at 5
B = 0.05Pm with respect to the minimum pattern size Pm,
It is preferable that c = 0.10 Pm and d = 0.75 Pm. For example, if the line-and-space pattern in FIG. 6A is a = Pm / 2, the line width a is about the resolution limit, and b, c, and d are expressed using a.
0.1a, c = 0.2a and d = 1.5a. When a correction pattern as shown in FIG. 6B is used, even if a projection exposure apparatus employing a special illumination method as shown in FIG. (Negative resist), and a fine pattern having a line width on the order of the resolution limit transferred onto the wafer does not taper and is as designed. Here, an isolated edge refers to an edge in which no other pattern (pattern edge) exists in the background adjacent to an edge within the minimum pattern size Pm from the edge.

FIG. 7A shows an isolated pattern P0 (line width a) of a light-transmitting portion provided on the light-shielding portion ground. It is assumed that the isolated pattern P0 and the periodic pattern of FIG. 6A are mixed in the same reticle. When the pattern P0 is exposed by the projection exposure apparatus shown in FIG. 1, the line width becomes narrow and the taper occurs as shown in FIG. 7C. Such an isolated pattern P0 is corrected as shown in FIG. First, the line width of the end portion (length d) of the pattern element P0 is set to c + in the X direction.
e, and the line width of the terminal portion is set to a + c + e. At this time, the line width a of the pattern element P0 is symmetrically increased by (c + e) / 2. The reason why the line width is increased by c + e is that both the correction for increasing the line width by e and the correction for increasing only the end portion by c are performed simultaneously. For the central part (other than the end part) of the pattern element, the line width is increased by e in the X direction and a + e
And At this time, the line width a of the pattern element P0 is symmetrically increased by e / 2.

Here, the dimensions of c and d are c = 0.1 Pm and d = 0.75 Pm, as in FIG.
Is about e = 0.1 Pm. By exposing the pattern of FIG. 7B with the apparatus of FIG. 1, a resist image (negative resist) without deformation as shown in FIG. 7D can be obtained. Note that a resist image with no taper and a thin line width can be obtained unless the taper is corrected (the end portion is thickened by c) and the line width is not corrected (thick by e). This means that the resolution of the isolated pattern is improved.

FIG. 8A is an example of a pattern in which a line-and-space pattern (periodic pattern) and an isolated pattern coexist, and the periodic patterns (P5, P6, P
7, P8: Both the line width and interval a) and the isolated pattern P7a (line width a) of the light-transmitting portion are provided in the light-shielding portion ground. FIG.
When the pattern of (A) is exposed using the exposure apparatus of FIG. 1, the resist image is deformed as shown in FIG. 8 (C).

FIGS. 6B and 7 show such a pattern.
The correction is performed as shown in FIG. 8B by the same method as the correction in FIG. First, the pattern elements P6 and P7 in the line and space pattern section are the pattern elements P2 and P7 in FIG.
3, and the pattern elements P5 and P9 are changed as shown in FIG.
The correction is made in the same manner as the pattern elements P1 and P2 in (A). Isolated pattern element P protruding from line and space
The lower end of the pattern element P7 including the pattern element 7a is shown in FIG.
The correction is made in the same manner as in the end of the pattern element P2 in FIG.
The isolated pattern element P7a is corrected in the same manner as the upper part of FIG. The line width of the isolated pattern element P7a is
E from the position f away from the line and space
Just fat. At this time, f is from 0.25 Pm to 0.5.
It is good to be about Pm. By exposing the pattern of FIG. 8B with the apparatus of FIG. 1, a resist image (negative resist) without deformation is obtained as shown in FIG. 8D.

All of the above-described patterns were patterns of light-transmitting portions on the ground of the light-shielding portions. This is because it is assumed that a resist image is formed on a wafer by using a negative resist. Conversely, even if a pattern of the light-shielding portion is used on the ground of the light-transmitting portion, the pattern may be quantitatively corrected in the same manner as described above. In this case, if a positive resist is used, a resist image with a line left on the wafer can be formed.

Here, the reticle is modified as described above using the reticle of the transmission portion pattern in the light shielding portion, and if a negative resist is used, the taper at the end portion in the longitudinal direction of the pattern is reduced.
In addition, both the end portions of the periodic pattern and the thinning of the line width in the isolated pattern were improved, and very good results were obtained. Therefore, with regard to the formation of a line-remaining pattern that is often included in memories and the like, the use of a negative resist is particularly effective in the exposure method according to the present invention. Conversely, when forming a circuit pattern containing a large number of line drawing patterns, the use of a positive resist is effective. This is a line leaving pattern,
The line-drawing pattern is caused by reversing the brightness of an optical image.

In order to correct the line width of the isolated pattern to a large value, the periodic pattern and the reticle in which the isolated pattern are mixed are exposed to a resist pattern of a desired size with a constant exposure amount. This is for transferring the pattern. Therefore, as described in the operation section, the same effect can be obtained by reducing the line width of the periodic pattern portion, contrary to the above-described correction.

That is, contrary to the conventional pattern correction, the line width of a pattern without any other pattern in the vicinity (isolated pattern) is not changed, and the pattern with another pattern in the vicinity (close pattern) is not changed. (A pattern having an edge). Similarly, in the vicinity of the longitudinal end of the linear pattern, the line width is not changed, and the center portion (except for the vicinity of the end) is not changed.
Modify the line width of. In this case, particularly when the pattern is a transmission portion pattern, the image contrast of the periodic pattern may be further improved. It is also suitable for forming a line pattern using a negative resist as described above. FIGS. 9A and 9B are diagrams showing patterns corrected according to this correction method.
(A) and (B) are examples in which the line width of each of the patterns in FIGS. 6 (A) and 7 (A) is corrected to be reduced. The correction amount is the same as in the case where the line width is increased. Note that when the correction is performed in the direction in which the line width is reduced, the optimum exposure amount is different from the correction in the direction in which the line width is reduced. In the case of the present embodiment, in both cases where the line width is increased or decreased, the exposure is performed at an exposure amount where the resist line width becomes a.

The square isolated pattern can be corrected by increasing the area around the same by 0.05 Pm as in the case of the line width correction amount in FIG. 7B. In particular, when a plurality of square isolated patterns of light-transmitting portions having different sizes from each other are formed on the light-shielding portion ground, correction is made so that small squares are enlarged in order to maintain mask linearity. In the case of a pattern as shown in FIG. 5C, the outermost edge surrounding the entire pattern may be extended outward by about 0.05 Pm (Px, Py).

FIG. 10 shows a second embodiment of reticle pattern correction. FIG. 10 (A) shows an L-shaped pattern element P10 (line width a) of a light-transmitting portion provided on a light-shielding portion ground. is there. The modification of the end of the L-shaped pattern element is the same as the modification of the end of the pattern element P0 in FIG. The correction of the line width is almost the same as the correction of the line width of the pattern element P0 in FIG. However, K1 outside the L-shaped corner portion
Regarding the above, the corner portion K1 is not corrected simply by increasing the line width in the short direction of the L-shaped pattern. Thus, for example, the correction of the L-shaped outer edge E15 is performed by expanding the edge E15 that has been lengthened by expanding the edge E16 so that the corrected edges are continuous.

For the inside K2 of the L-shaped corner,
When the thickness is increased similarly to the other portions, the inside of the resist pattern bulges as shown by the dotted line in FIG. 9C, and the fidelity of the image is reduced. Therefore, the line width of the corner portion is preferably set to be smaller than that of the other portion (in this embodiment, the edge of the inner portion is not corrected, and the edge of the outer portion is corrected to increase the line width of the corner portion. ing.). In the present embodiment, the line width of the corner portion is increased by b and is approximately a + b. The length g for reducing the amount of fat is 0.25 Pm to 0.5.
It is good to be about Pm. Thus, the length g of the corner portion
The reason why the line width thickening amount is made smaller is that the edge between the corners near the inner portion K2 has the minimum pitch Pm in the adjacent ground.
This is because the edges are not separated by more than a degree, and the edges are considered to be not isolated.

By exposing the pattern of FIG. 10B with the apparatus of FIG. 1, a resist image (negative resist) without deformation as shown by a solid line in FIG. 10C is obtained. FIG.
FIG. 11A shows a third embodiment of reticle pattern correction, and FIG. 11A shows a pattern element P11 (line width a) of a light-transmitting portion provided in a light-shielding portion. The modification of the pattern element P11 is performed in the same manner as in FIG. 10B for the end portion, the center portion (other than the end portion), and the inside portion of the corner. Therefore, FIG.
The pattern shape may be corrected as shown in FIG.

The modification shown in FIGS. 10B and 11B is also applicable to the case where a light-shielding portion pattern is formed on a light-transmitting portion, assuming that a line-remaining pattern is transferred to a positive resist. Is effective. Also, in the second and third embodiments, the line width may be modified in the same manner as in the first embodiment to reduce the line width.

Although not shown in FIG. 6, FIG. 7, FIG. 8, FIG. 10, and FIG. 11, the line width in the longitudinal direction of the end portion of each pattern element is also very small (b = 0.05 Pm
Degree) It is desirable to make it fat. For example, when correcting the pattern element P1 of FIG. 6A, the edges E9 and E10 whose end portions are corrected and whose length in the X direction is a + b + c are changed to Y.
What is necessary is just to extend by b in the direction.

[0065]

As described above, according to the present invention, the resolution and the depth of focus can be greatly improved, and the periodic pattern,
Good image quality can be obtained for all circuit patterns such as isolated patterns. According to one aspect of the present invention, a periodic pattern is formed by a mask including only a normal light-transmitting portion and a light-shielding portion partially formed with a correction pattern.
Good image quality can be obtained for all patterns such as isolated patterns, and a mask that is easier to manufacture, inspect, correct, and clean than a phase shift mask can be obtained.

Further, according to another aspect of the present invention, by illuminating the mask obliquely using a mask in which a part of a pattern composed of only a normal light-transmitting portion and a light-shielding portion is corrected, a higher resolution than in the past can be achieved. A large depth of focus can be obtained, and good image quality can be obtained for all patterns such as periodic patterns and isolated patterns. In particular, when a negative resist is used, a line-remaining pattern having a good resist image profile can be formed.

[Brief description of the drawings]

FIG. 1 is a diagram showing a schematic configuration of a projection exposure apparatus suitable for an exposure method according to an embodiment of the present invention;

FIG. 2 is a diagram showing a light beam splitting member and a movable portion according to one embodiment of the present invention as viewed from an optical axis direction;

FIG. 3 is a diagram illustrating a light beam splitting member and a movable portion according to an embodiment of the present invention when viewed from an optical axis direction;

FIG. 4 is a diagram schematically showing an optical path from a Fourier transform plane of a mask to a projection optical system in the apparatus of FIG. 1,

5A and 5C are plan views showing an example of a reticle pattern formed on a mask, and FIGS. 5B and 5D are pupil conjugate planes corresponding to FIGS. 5A and 5C, respectively. FIG.

6A is a diagram illustrating a periodic pattern, FIG. 6B is a diagram illustrating a correction pattern of FIG. 6A according to an embodiment of the present invention,
(C) is a diagram showing a resist image when the pattern of (A) is exposed by the apparatus of FIG. 1; (D) is a diagram showing a resist image when the pattern of (B) is exposed by the apparatus of FIG. 1;

7A is a diagram showing an isolated pattern, FIG. 7B is a diagram showing a correction pattern of FIG. 7A according to an embodiment of the present invention, and FIG.
FIG. 3A is a view showing a resist image when the pattern of FIG. 1A is exposed by the apparatus of FIG. 1; FIG. 3D is a view showing a resist image when the pattern of FIG. 1B is exposed by the apparatus of FIG.

8A is a diagram showing a pattern in which a periodic pattern and an isolated pattern coexist; FIG. 8B is a diagram showing a correction pattern in FIG. 8A according to an embodiment of the present invention; FIG. 3 is a view showing a resist image when the pattern is exposed by the apparatus of FIG. 1; FIG. 3D is a view showing a resist image when the pattern of (D) and (B) is exposed by the apparatus of FIG.

FIG. 9A shows a modification of the pattern correction of FIG . 6A.
FIG. 7B shows a modification of the pattern correction of FIG.
Figure,

FIG. 10A illustrates an L-shaped isolated pattern,
(B) is a diagram showing a correction pattern of (A) according to an embodiment of the present invention, (C) is a diagram showing a resist image when the pattern of (B) is exposed by the apparatus of FIG.

11A is a diagram showing an isolated pattern, FIG. 11B is a diagram showing a correction pattern of FIG. 11A according to an embodiment of the present invention,

12A is a diagram illustrating a reticle pattern in which a periodic pattern is formed by a light-transmitting portion on a light-shielding portion, FIG. 12B is a diagram illustrating a resist image in a cross section taken along line M1-M2 of FIG.
FIG. 5A is a view showing a resist image in an M3-M4 cross section of FIG.
(D) is a diagram showing a resist image when the pattern of (A) is exposed by the apparatus of FIG. 13; (E) is a diagram showing a reticle pattern in which a periodic pattern formed by a light-shielding portion is formed on a transparent portion;
FIG. 13 (F) shows a resist image in M1-M2 section, FIG. 13 (G) shows a resist image in M3-M4 section in (E), and FIG. A diagram showing a resist image when exposed,

FIG. 13 illustrates the principle of the present invention;

FIG. 14 is a diagram showing a schematic configuration of a conventional projection exposure apparatus.

[Explanation of symbols]

 7 light beam splitting member 11 reticle 12 reticle pattern 13 projection optical system 15 Fourier transform plane 18 pupil plane 19 wafer AX optical axis

──────────────────────────────────────────────────続 き Continuation of the front page (51) Int.Cl. ⁷ Identification symbol FI H01L 21/027 H01L 21/30 502P 528 (56) References JP-A-3-36549 (JP, A) JP-A-1-188857 (JP, A) JP-A-56-12615 (JP, A) "1987 SYMPOSIUM ON VLSI TECHNOLOGY" DIGEST OF TECHNICAL PAPERS, IEEE CAT. NO. 87, 1987, pages 13, 14

Claims (35)

(57) [Claims] 1. A mask used in an exposure apparatus for transferring a pattern onto a photosensitive substrate, wherein the pattern has a pair of edges in which at least one of a plurality of pattern elements is parallel, and the pair of edges is another one. At least a portion includes an isolated edge whose distance from the pattern element exceeds a predetermined value, and the at least one pattern element has a width defined by the pair of edges larger than a design value at the isolated edge. And a width defined by the pair of edges is relatively larger near the end than at the center. 2. The mask according to claim 1, wherein only one of said pair of edges is said isolated edge, and said at least one pattern element is extended with said one edge. 3. The mask according to claim 1, wherein said pair of edges are each said isolated edge, and said at least one pattern element is such that said pair of edges is respectively extended. 4. The method according to claim 1, wherein the at least one pattern element has a rectangular shape having a pair of edges orthogonal to the pair of edges and serving as the isolated edge, and each of the four edges is expanded. The mask according to claim 1, wherein: 5. The plurality of pattern elements each have a pair of edges extending along a first direction, and are periodically arranged in a second direction orthogonal to the first direction. 2. The mask according to claim 1, wherein the outer edges are extended so that the width is larger than the design value in each of the pattern elements at both ends. 3. 6. A mask used in an exposure apparatus for transferring a pattern onto a photosensitive substrate, wherein the pattern has a pair of edges in which at least one of a plurality of pattern elements is parallel to each other, and the pair of edges is respectively In some cases, the at least one pattern element has a width defined by the pair of edges, which is a distance from another pattern element exceeding a predetermined value, and the width defined by the pair of edges is greater than the remaining edges. A mask which is relatively thick, and a width defined by the pair of edges is relatively thick near a terminal end than at a central portion. 7. The mask according to claim 1, wherein the predetermined value is determined according to a resolution limit of the exposure apparatus. 8. A mask used in an exposure apparatus for transferring a pattern onto a photosensitive substrate, wherein the pattern has a pair of edges in which at least three of a plurality of pattern elements each extend along a first direction, and The at least three pattern elements are periodically arranged in a second direction orthogonal to the first direction, and the width defined by the pair of edges is relatively larger at the both end pattern elements than at the remaining pattern elements. And a width of each of the at least three pattern elements defined by the pair of edges is relatively larger near a terminal end than at a central portion. 9. The mask according to claim 8, wherein the width of each of the pattern elements at both ends is larger than a design value. 10. The mask according to claim 8, wherein the width of the remaining pattern elements is smaller than a design value. 11. Each of the pattern elements arranged in the second direction includes a line pattern having the first direction as a longitudinal direction,
Alternatively, the mask according to any one of claims 5, 8 to 10, wherein the mask is a square pattern. 12. A mask used in an exposure apparatus for transferring a pattern onto a photosensitive substrate, wherein the pattern has a line width defined by a pair of edges in which at least one of a plurality of pattern elements is parallel, and the pair of edges includes At least a part includes a straight line part whose end is defined by an edge orthogonal to the straight line part,
2. The mask according to claim 1, wherein the line width is relatively larger near the end than at the center. 13. The mask according to claim 12, wherein the pair of edges of the straight portion is expanded so that the line width becomes larger than a design value near the terminal end. 14. The mask according to claim 12, wherein the pair of edges of the linear portion is reduced so that the line width at the central portion is smaller than a design value. 15. The method according to claim 1, wherein the design value is a line width on the mask of a desired line width of the pattern transferred on the photosensitive substrate.
15. The mask according to any one of items 3 and 14. 16. The exposure apparatus according to claim 1, wherein the exposure apparatus defines the illumination light within a predetermined area off an optical axis on a pupil plane of an illumination optical system that irradiates the pattern with illumination light.
The mask according to any one of the preceding claims. 17. The illumination light for irradiating the mask according to claim 1 with illumination light and exposing a photosensitive substrate with the illumination light via a projection optical system. An exposure method comprising: defining the illumination light in a predetermined area off the optical axis on a pupil plane of an illumination optical system that irradiates the illumination light onto the mask. 18. An exposure method according to claim 17 , wherein a negative tone type photoresist is used as a photosensitive agent for said substrate. 19. An exposure apparatus comprising: an illumination optical system for irradiating the mask according to claim 1 with illumination light; and a projection optical system for projecting the illumination light onto a photosensitive substrate. A method for manufacturing a circuit element, comprising a lithography step of transferring a pattern of the mask onto the substrate using an apparatus. 20. The illumination light system according to claim 15, wherein the illumination light is defined in a predetermined area off the optical axis on a pupil plane of the illumination optical system.
20. The method for manufacturing a circuit element according to 19 . 21. The pattern has a pattern element whose longitudinal direction is a first direction, and the predetermined region intersects with an optical axis on a pupil plane of the illumination optical system and extends along the first direction. 21. The method according to claim 20 , further comprising a region defined by a first axis defined. 22. The pattern has a pattern element whose longitudinal direction is a second direction orthogonal to the first direction, and the predetermined region is defined by the first axis and the optical axis of the illumination optical system. 22. The circuit element manufacturing method according to claim 21 , comprising a region orthogonal to one axis and separated by a second axis defined along the second direction. 23. The pattern has a pattern element having a longitudinal direction in each of a first direction and a second direction, wherein the predetermined region has a first distance from the optical axis in the first direction that is substantially equal, and 23. The circuit element manufacturing method according to claim 20 , comprising a plurality of regions having a second distance from the optical axis in the second direction that is substantially equal. 24. The plurality of regions, wherein the first distance is determined according to the fineness of the pattern element in the first direction, and the first distance is determined according to the fineness of the pattern element in the second direction. The method according to claim 23 , wherein two distances are determined. 25. The pattern has a pattern element whose longitudinal direction is a first direction, and the predetermined region is substantially parallel to the first direction, and the light is emitted in a second direction orthogonal to the first direction. A plurality of regions arranged on a pair of first line segments separated from an axis by a distance according to a degree of fineness of the pattern element in the second direction are included.
25. The circuit element manufacturing method according to any one of 20 to 24 . 26. The pattern having a pattern element having the second direction as a longitudinal direction, wherein the plurality of regions are substantially parallel to the second direction, and the pattern is formed from the optical axis with respect to the first direction. The element is disposed at an intersection of a pair of second line segments separated by a distance according to the degree of fineness of the element in the first direction and the pair of first line segments.
26. The circuit element manufacturing method according to 25 . 27. The predetermined area includes a plurality of areas decentered from an optical axis on a pupil plane of the illumination optical system, and two different orders generated from the pattern by irradiation of light emitted from the respective areas. As the diffracted light passes through a position on the pupil plane of the projection optical system where the distance from the optical axis is substantially equal,
The position of each of the areas is determined.
The method for manufacturing a circuit element according to any one of claims 20 to 26 . 28. The circuit element manufacturing method according to claim 27 , wherein said two diffracted lights are one of ± n order diffracted lights and a 0 order diffracted light. 29. The predetermined area includes a plurality of areas decentered from an optical axis on a pupil plane of the illumination optical system, wherein an incident angle of light emitted from each area to the mask is ψ, the diffraction angle of the generated ± n order diffracted light theta, the number of mask-side numerical aperture of the projection optical system and NA _R, when the sin (θ-ψ) = NA R the relationship is satisfied in one of the ± n order diffracted light The method according to any one of claims 20 to 28 , wherein a resolution limit of the projection optical system is set. 30. The method according to claim 29 , wherein said one diffracted light is substantially symmetric with respect to an optical axis of said projection optical system with respect to a zero-order diffracted light generated from said pattern. 31. The predetermined region includes a plurality of regions decentered from an optical axis on a pupil plane of the illumination optical system, wherein a wavelength of the illumination light is λ, a pitch of the pattern is P, and light is emitted from each of the regions. The incident angle ψ of the light to be incident on the mask is sinψ =
The method according to any one of claims 20 to 30 , wherein the position of each of the regions is determined so as to satisfy a relationship of λ / 2P. 32. The predetermined area includes a plurality of areas decentered from an optical axis on a pupil plane of the illumination optical system, wherein the angle of incidence of light emitted from each area on the mask is ψ, claims the wavelength lambda, the numerical aperture of the mask side of the projection optical system and NA _R, the minimum pitch of the transferable pattern on the substrate is characterized in that it is a _{λ / (NA R + sinψ)} 20 32. The method for manufacturing a circuit element according to any one of items 31 to 31 . 33. The wavelength of the illumination light lambda, the numerical aperture of the mask side of the projection optical system and NA _R, the pattern is characterized in that the pitch has a small pattern elements than lambda / NA _R The method for manufacturing a circuit element according to claim 20 . 34. The pattern has pattern elements each having a first direction and a second direction as longitudinal directions, and the predetermined area includes a plurality of areas decentered from an optical axis on a pupil plane of the illumination optical system, One of the 0th-order diffracted light generated from the pattern by irradiation of light emitted from each of the regions, and the higher-order diffracted light distributed in the first direction around the 0th-order diffracted light.
And the position of each of the regions such that one of the higher-order diffracted lights distributed in the second direction around the zero-order diffracted light is distributed at substantially the same distance from the optical axis on the pupil plane of the projection optical system. The method according to any one of claims 20 to 33 , wherein is determined. 35. A claim, wherein the higher-order diffracted light distributed to the first and second directions is the order from each other is equal to 34
3. The method for producing a circuit element according to claim 1.